Recent crystallographic studies have revealed the detailed structure of several antibody/protein antigen complexes. What remains to be understood is the relationship between geometric proximity of potentially=interacting groups, as observed in the complex, and the affinity and selectivity of molecular recognition of the antigen by the antibody. The relationship between structure and molecular recognition is complicated by dynamical motions of residues in the contact region and electrostatic interactions in the complex dielectric environment of the protein/protein interface. It is desireable, therefore, to take a molecular-scale, thermodynamic approach to the structural basis of antibody/antigen affinity. We will use one of the few antibody/protein complexes of known structure, that formed by hen egg lysozyme with the FAb HyHEL-5, for a detailed study of the structural and thermodynamic basis of antibody binding affinity. Directed mutagenesis of the antigen by us, and of the antibody by our collaborator Dr. S. Smith-Gill of the N.I.H., will allow detailed testing of the contributions of specific residues and contacts to the affinity and specificity of binding. These contributions will be characterized in terms of their effects on binding affinity, and also their effects on the enthalpy of binding using titration microcalorimetry. This combined approach will help to reveal the basic thermodynamic driving forces for molecular recognition in this complex. Free energy simulation will be used to predict the relative affinities and enthalpies of binding of the various mutant complexes, and will also provide insight into the role of protein dynamics and conformational changes in the recognition process. Comparison of the simulation results with experiment will also assist in the further development of the theoretical methods. The significant of the proposed work lies in its potential to advance our understanding of the fundamental basis of molecular recognition. The proposed work represents the first integrated study of a single protein/protein complex of known structure applying the methods of directed mutagenesis, calorimetry, and molecular simulation in a synergistic way. For this reason, we believe that the proposed work has the potential to produce new, general insights into the basis of molecular recognition in antigen/antibody complexes.